Category: Digital Health
Room: Indigo Ballroom
Objective: Most mobile health apps lack scientific evidence. The One Drop mobile app offers tracking of diabetes self-care behaviors and blood glucose (BG) values manually or automatically via Bluetooth-enabled meters, CGMs, and other health apps. The app includes a robust food library, medication scheduler, recipes, health tips, user polls, and community support (‘likes’, stickers, and data sharing). Users specify BG and self-care goals, and receive data-driven insights to draw connections between behaviors, goals, self-reported hemoglobin A1c and weight. In July 2016, we assessed A1c change among One Drop users self-reporting a diabetes diagnosis.
Method: We queried data on ~50,000 users, identifying those users with two A1c values at least 60 and no more than 365 days apart. A two-sided paired t-test assessed changes from first to last A1c. Two-sided independent t-tests assessed gender and diabetes type differences on the A1c change score, and Spearman’s correlation coefficients assessed relationships between diabetes duration, tracking data, and A1c change.
Results: The sample included 133 users (63 males), including 41 with type 1 (T1D) and 88 with type 2 diabetes (T2D). Average diabetes duration was 12.4 ± 10.8 years with 44.2% using insulin and 50.4% using the One Drop medication scheduler. The average time interval between A1c values was 146.1 ± 76.7 days, with 3,225 ± 5,780 (range: 2-37,838) tracking entries during that period. A1c improved significatly from initial (7.8% ± 2.1%) to final period (7.1% ± 1.3%), two-sided t=4.2, p < .001, and the change in mean A1c was similar irrespective of gender or diabetes type. However, subgroup analyses by diabetes type revealed substantial A1c improvement among people with T2D (7.7% ± 2.2% vs. 6.9% ± 1.2%; t=4.0, p < .001), with a trend of improvement among people with T1D (8.0% ± 2.0% vs. 7.5% ± 1.3%; t=1.7, p=0.09). Changes in A1c were not associated with duration of diabetes or tracking intensity.
Conclusion: People with diabetes using the One Drop health app reported a nearly 0.70% reduction in A1c during 2-12 months of using it. Mean A1c change did not differ by gender or diabetes type. Ongoing analyses of user data combined with results from a third-party trial will provide further evidence of One Drop’s self-care and clinical value.
Brian Huddleston– VP of Engineering, Informed Data Systems Inc., Austin, Texas
Chandra Osborn– VP of Health and Behavioral Informatics, One Drop, Informed Data Systems Inc., Nashville, Tennessee
David Rodbard– President, Biomedical Informatics Consultants LLC, Potomac, Maryland
Jeff Dachis– Founder & CEO, Informed Data Systems Inc., New York, New York
VP of Engineering
Informed Data Systems Inc.
VP of Health and Behavioral Informatics
One Drop, Informed Data Systems Inc.
Biomedical Informatics Consultants LLC
Founder & CEO
Informed Data Systems Inc.
New York, New York